Error: during training on own datasets

[pp3993]

Hi Canjie-Luo
I referred the issue #44
I am training the model on my own dataset with the following alphabet list:

alphabet='0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ()-.,>:# /'
Example datasets used for training.

• ELGRENZVYULJI :2))7 ,91 < 9oR1#t4h80Njw

• Kgpyhzyszobme

• YBOBTOVJL 6<<'830 :)-4<9 .17

I have following doubts:

1. As my data is having maximum width upto 565 pixels, Do i need to make any changes in targetW or ImageW parameter?

2. Is it works for long text predictions?

Other than this I got the following error while running the main.py code.

To solve the below error, I used the following solution as given in stackoverlow, but did not work for me.
import sys
sys.setrecursionlimit(3000)

RequestsDependencyWarning)
Namespace(BidirDecoder=True, MORAN='', adadelta=True, adam=False, alphabet='0;1;2;3;4;5;6;7;8;9;a;b;c;d;e;f;g;h;i;j;k;l;m;n;o;p;q;r;s;t;u;v;w;x;y;z;A;B;C;D;E;F;G;H;I;J;K;L;M;N;O;P;Q;R;S;T;U;V;W;X;Y;Z;(;);-;.;,;>;:;#; ;/;$', batchSize=32, beta1=0.5, cuda=True, displayInterval=100, experiment='output/', imgH=64, imgW=200, lr=1.0, n_test_disp=10, ngpu=1, nh=256, niter=100, saveInterval=40000, sep=';', sgd=False, targetH=32, targetW=100, train_nips='/home/payal/FRSLABS/MORAN_v2-master/demo/mdbiit5k/', valInterval=100, valroot='/home/payal/FRSLABS/MORAN_v2-master/demo/mdbiit5k/', workers=1)
mkdir: cannot create directory ‘output/’: File exists
Random Seed:  1818
Start val
Traceback (most recent call last):
  File "main.py", line 245, in <module>
    acc_tmp = val(test_dataset, criterion)
  File "main.py", line 159, in val
    t, l = converter.encode(cpu_texts, scanned=True)
  File "/home/payal/FRSLABS/MORAN_v2-master/tools/utils.py", line 86, in encode
    text, _ = self.encode(text)
  File "/home/payal/FRSLABS/MORAN_v2-master/tools/utils.py", line 86, in encode
    text, _ = self.encode(text)
  File "/home/payal/FRSLABS/MORAN_v2-master/tools/utils.py", line 86, in encode
    text, _ = self.encode(text)
  [Previous line repeated 985 more times]
  File "/home/payal/FRSLABS/MORAN_v2-master/tools/utils.py", line 82, in encode
    elif isinstance(text, collections.Iterable):
  File "/home/payal/anaconda3/lib/python3.6/abc.py", line 184, in __instancecheck__
    if subclass in cls._abc_cache:
  File "/home/payal/anaconda3/lib/python3.6/_weakrefset.py", line 75, in __contains__
    return wr in self.data
RecursionError: maximum recursion depth exceeded in comparison
Exception ignored in: <bound method DataLoaderIter.__del__ of <torch.utils.data.dataloader.DataLoaderIter object at 0x7fb9dc3e04e0>>
Traceback (most recent call last):
  File "/home/payal/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 333, in __del__
    self._shutdown_workers()
  File "/home/payal/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 319, in _shutdown_workers
    self.data_queue.get()
  File "/home/payal/anaconda3/lib/python3.6/multiprocessing/queues.py", line 337, in get
    return _ForkingPickler.loads(res)
  File "/home/payal/anaconda3/lib/python3.6/site-packages/torch/multiprocessing/reductions.py", line 70, in rebuild_storage_fd
    fd = df.detach()
  File "/home/payal/anaconda3/lib/python3.6/multiprocessing/resource_sharer.py", line 57, in detach
    with _resource_sharer.get_connection(self._id) as conn:
  File "/home/payal/anaconda3/lib/python3.6/multiprocessing/resource_sharer.py", line 87, in get_connection
    c = Client(address, authkey=process.current_process().authkey)
  File "/home/payal/anaconda3/lib/python3.6/multiprocessing/connection.py", line 487, in Client
    c = SocketClient(address)
  File "/home/payal/anaconda3/lib/python3.6/multiprocessing/connection.py", line 614, in SocketClient
    s.connect(address)
ConnectionRefusedError: [Errno 111] Connection refused
Exception ignored in: <bound method DataLoaderIter.__del__ of <torch.utils.data.dataloader.DataLoaderIter object at 0x7fb9dc3b83c8>>
Traceback (most recent call last):
  File "/home/payal/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 333, in __del__
  File "/home/payal/anaconda3/lib/python3.6/site-packages/torch/utils/data/dataloader.py", line 319, in _shutdown_workers
  File "/home/payal/anaconda3/lib/python3.6/multiprocessing/queues.py", line 337, in get
ImportError: sys.meta_path is None, Python is likely shutting down

Please help!

[Canjie-Luo]

Yes, you should enlarge the image width, both input width and target width. Please make sure the data is correctly prepared for the network:
File "/home/payal/FRSLABS/MORAN_v2-master/tools/utils.py", line 82, in encode elif isinstance(text, collections.Iterable)
Actually, attention mechanism works well on short text. I am not sure that it is fine for your task.